How do I get jq to take json like this:
{
"host1": { "ip": "10.1.2.3" },
"host2": { "ip": "10.1.2.2" },
"host3": { "ip": "10.1.18.1" }
}
and generate this output:
host1, 10.1.2.3
host2, 10.1.2.2
host3, 10.1.18.1
I'm not interested in the formatting, I just can't figure out how to access the key name and value.
To get the top-level keys as a stream, you can use the built-in function keys[]
. So one solution to your particular problem would be:
jq -r 'keys[] as $k | "\($k), \(.[$k] | .ip)"'
keys
produces the key names in sorted order; if you want them in the original order, use keys_unsorted
.
Another alternative, which produces keys in the original order, is:
jq -r 'to_entries[] | "\(.key), \(.value | .ip)"'
The @csv and @tsv filters might also be worth considering here, e.g.
jq -r 'to_entries[] | [.key, .value.ip] | @tsv'
produces:
host1 10.1.2.3
host2 10.1.2.2
host3 10.1.18.1
If the keys of interest are embedded as in the following example, the jq filter would have to be modified along the lines shown.
Input:
{
"myhosts": {
"host1": { "ip": "10.1.2.3" },
"host2": { "ip": "10.1.2.2" },
"host3": { "ip": "10.1.18.1" }
}
}
Modification:
jq -r '.myhosts | keys[] as $k | "\($k), \(.[$k] | .ip)"'